Experimental Investigation of the Air Release in Hydraulic Reservoirs

Transcription

1 Group M - Hydraulic Components Paper M Experimental Investigation of the Air Release in Hydraulic Reservoirs Marco Longhitano, M. Sc. RWTH Aachen University, Institute for Fluid Power Drives and Controls (IFAS), Aachen, Germany, Marco.Longhitano@ifas.rwth-aachen.de Alessandro Protase, B. Sc. RWTH Aachen University, Institute for Fluid Power Drives and Controls (IFAS), Aachen, Germany Professor Dr.-Ing. Hubertus Murrenhoff RWTH Aachen University, Institute for Fluid Power Drives and Controls (IFAS), Aachen, Germany Abstract Air contamination strongly decreases the efficiency of fluid power systems and when the allowable limits are exceeded, the performance of the system deteriorates. The hydraulic reservoir performs the function of releasing the entrained air of the hydraulic system to the surroundings. In recent years, the reservoir design has become an important task in the design of the hydraulic system due to space restrictions forcing the use of small sized reservoirs. Despite this fact, experimental results on an air release are not available. In this paper, an experimental investigation of the air release in hydraulic reservoirs is presented. A test apparatus using an optical method as well as the post-processing of the results is described. These are given in terms of an air release rate for different reservoir designs over a wide range of oil flow rates and air loads. The current study is a significant step forward in the design of fluid power systems, as it provides an experimental procedure to measure the air release in the hydraulic reservoir as well as its quantitative analysis. KEYWORDS: Reservoir, air release, optical method, experimental analysis 1. Introduction In the last decades, in response to growing concerns over the environmental impacts of mobile machinery and their costs, manufacturers are being forced to develop more energy efficient hydraulic systems and to decrease the used fluid volume by decreasing the reservoir size, while maintaining the lowest possible amount of air

2 598 10th International Fluid Power Conference Dresden 2016 drawn into the pump. Another aspect strongly influencing the size and the design of the reservoirs is the additional space requirement for components for after-treatment of the exhaust gas. As a result, much effort has been put into studying the air release in hydraulic reservoirs /1/. Entrained air forms inside hydraulic components due to cavitation /2/ or malfunctioning seals. This leads to problems in hydraulic systems. First of all, it changes fluid properties such as density, viscosity and bulk modulus /3,4/. This results in an increased oil compressibility, which causes pressure losses and changes the stiffness of the system. Phenomena such as the diesel effect /5/ and cavitation erosion may occur and damage the components of the system. Finally, a high amount of air in oil produces more noise. The problems mentioned diminish the whole efficiency of the system and the endurance of the fluid. They make the design, the simulation, the manufacturing and maintenance of fluid power systems more difficult and expensive. In order to prevent the above mentioned disadvantages, the air release in a hydraulic system has been studied under the simulative and experimental point of view. An overview of the different methods to measure the air amount in oil has been discussed by Schrank et al. in /6/. The most promising ones involve optical techniques such the PIV (Particle Imaging Velocimetry) /7/ and the shadowgraph method. The air bubble behaviour in mineral oil has been experimentally investigated in terms of bubble velocity and coalescence /8/. The first measurement of the air release efficiency in the tank geometries can be found in /9/. Recently, Wohlers published the measurements of the air release in an hydraulic system as function of time /10/. Different CFD methods are employed to simulate the behaviour of an oil-air mixture. A novel application of such models is the simulation of a hydraulic reservoir conducted by Ti et al. and Untch et al. in /11/ and /12/ respectively. The hydraulic reservoir performs the function of releasing the entrained air to the surroundings. A sub-optimal design of the reservoir leads to a poor air release consequently decreasing the efficiency of the system and aggravates the above mentioned problems. Unfortunately, the air release capability of a hydraulic reservoir as a function of its working conditions has not been investigated. It leads to the problems in the validation of the simulation codes and in the design of the reservoir. This paper presents an experimental study of the air release in hydraulic reservoirs. After a theoretical explanation of the air release process, experimental apparatus developed at the Institute for Fluid Power Drives and Controls at RWTH Aachen University using the shadowgraph method is presented. The test results are shown in terms of an air release rate for different air loads and as a function of various oil flow

3 Group M - Hydraulic Components Paper M rates. The influence of the internal reservoir design has been investigated as well. Finally, a study of the bubble diameter distribution is shown. 2. Air release process Entrained air flows into the reservoir in bubble form. The different forces acting on a bubble moving upwards in a fluid result in a terminal velocity of the bubble. It can be written as in Eqn. (1) for the so-called viscous dominant regime /8/. (1) The velocity is proportional to the square of the bubble diameter and to the air and oil properties. Eqn. (1) shows an inverse proportionality of the velocity to the Reynolds number of the bubble and to the drag coefficient. The latter describes the resistance, that the bubble experiences, due to the presence of the fluid and varies with temperature and the oil class /8/. Finally, the parameter refers to fixed geometrical constants. As shown in Figure 1, the bubbles enter in a reservoir from the inlet with the oil flow rate. The air load is described by the volume air fraction defined in Eqn. (2). (2) The bubble pattern is defined by means of the composition between the rise velocity expressed in Eqn. (1) and the oil velocity acting on the bubble. The latter depends on the oil flow rate, the volume of oil and the internal design of the reservoir (separating plates, sieves and baffles). Figure 1: Air release process in hydraulic reservoirs The bubbles rising with higher velocity can be released through the free surface of the oil ( ), while the rest flows into the outlet ( ). The air release rate quantifies the

4 600 10th International Fluid Power Conference Dresden 2016 capability of a hydraulic geometry to release the entrained air and it is defined as in Eqn. (3). % (3) 3. Experimental apparatus and measurements The experiments are performed in the IFAS laboratory. A shadowgraph method is used to record the images at the outlet of the test reservoir. The pictures are then postprocessed to measure the air content Test bench and shadowgraph method The circuit of the test bench is illustrated in Figure 2. Its goal is to measure the air release rate of the hydraulic reservoirs. In the first subsystem a variable displacement pump draws an oil flow rate between and from a main tank of capacity. The flow rate is regulated by means of a flow rate sensor. As can be seen in Figure 2, the oil then flows through the chamber of the second subsystem. This chamber is connected with a pneumatic system, which injects a defined air flow rate through porous elements. Figure 2: Circuit of the test bench and optical system The air and oil mixture flows in the test reservoir illustrated in the third subsystem. A part of the injected air is released inside the tank to the surroundings, while the rest is drawn into the pump as illustrated in the fifth subsystem. The flow rate of the second

5 Group M - Hydraulic Components Paper M pump is also regulated by means of an additional flow rate sensor. The pressure and the temperature up- and downstream of the test tank are measured. A part of the mixture is extracted before reaching the pump and flows into a plexiglas chamber placed in a by-pass parallel to the main line. The mixture before the extraction optically shows a homogenous and uniform distribution of the air in the fluid phase. This chamber has a rectangular section with a thickness of and is located between a high speed camera and a background light. The oil flows again in the main tank, which is designed to release the remaining part of the entrained air. Finally, the circuit illustrated in the sixth subsystem performs the filtration and the heat-exchange functions. The shadowgraph apparatus captures the images of the air bubbles at the outlet observing a flow exhibiting variations of the fluid density. It consists of a high speed camera equipped with a CMOS-Sensor, which can take up to fps at reduced resolution. A high-magnification lens system is mounted onto the camera with a resolution of. The main idea of this technique is that the light rays are refracted by the bubbles. With the help of the background light provided on the right side of the chamber, the camera on the left side is able to capture resulting light refraction as shadows. Additionally, a diffuser is located between the light source and the test tank to dampen and spread the light. The whole test setup and the specifications of the test reservoir are illustrated in Figure 3. The inclination of the test tank can be set either to or to. The test tank is manufactured according to the norm DIN24339 /13/ and has a max oil volume of. Figure 3: Specifications of the test reservoir

6 602 10th International Fluid Power Conference Dresden 2016 The average dwell time ranges between and at the maximum oil volume. The outlet port is located at the side of tank in order to avoid cavitation problems. The cover plate has five different inlet configurations and an air breather. Additionally, three inspection windows are placed at the tank walls monitoring the flow during the measurements. A separation plate can be mounted inside the tank between the inlet and the outlet ports Post-processing procedure After the oil and air flow rate are set and a steady-state is reached, the camera records the images of the air and the oil mixture. The pictures are post-processed using a script implemented in MATLAB. The goal of the script is to calculate the air release rate. This is accomplished in different steps: 1. Reading and filtering the image 2. Conversion of the image in binary and detection of the bubbles and their diameters 3. Calculation of the air fraction for each picture dividing the sum of the volume of the bubbles by the total volume of the view window 4. Statistical analysis: noise reduction and outlier elimination 5. Calculation of the bubble diameter distributions The post-processing script provides a result as measurement of the air fraction at the outlet for each measure. The air fraction at the inlet is calculated using the measured air and oil flow rate. Finally, the air release rate is calculated using Eqn. (3). 4. Results and discussion An ISO VG 46 mineral oil has been used to conduct all the tests. The test reservoir was filled with of oil. The different configurations of the test tank used for the measurements are given in Figure 4. Additionally, the boundary conditions of the tests in terms of air flow rate and oil flow rate are illustrated. In the first configuration, the inlet and the outlet pipes have the minimum distance. This has been tested for all the working conditions. In the second design, the inlet is located further from the outlet. The third and the fourth geometries differ from each other by the plate that separates the inlet and the outlet ports. The last three designs have been tested keeping either the oil flow rate equal to or the air flow rate equal to. The air flow rate data are reported at standard conditions of pressure and temperature. Both of them have not experienced any change during the measurements. This section presents the experimental results of the air release in the hydraulic reservoirs. The first

7 Group M - Hydraulic Components Paper M part shows the influence of the oil flow rate and the air load on the air release rate. Additionally, the experimental data of the bubble diameter distribution are given. The second part analyses the role of the internal design of the tank in the air release process. Figure 4: Tested configurations of the hydraulic reservoir 4.1. Variation of the working conditions The measured air release rate of the first tank design is illustrated in Figure 5a as a function of the injected oil flow rate for the different air flow rates. An increase of the oil flow rate leads to the decrease of the air release rate. This is due to the increase of the oil velocity inside the tank. This diminishes the time that the bubbles have to be released from the free surface of the fluid before entering the suction pipe. This trend can be seen in Figure 5b as well, where the bubble diameter distributions are plotted for different oil flow rates with a maximum bubble diameter of. According to the distribution of microbubbles of gas in water /14/, the smaller the bubbles are, the higher is the frequency. The three histograms in Figure 5b have a significant gap for all the classes due to the better air release at lower oil flow rate. The different curves at constant air flow rates in Figure 5a show that the air release rate sharply decreases with an increasing of the air load. This phenomenon attenuates at higher air flow rates. In order to observe such trend, the measured air release rate of the first reservoir design is illustrated in Figure 6a as a function of the injected air flow rate for the different oil flow rates. The air release rate sharply decreases with an increasing air load until for all the curves. Afterwards, the reduction diminishes and they start to have an asymptotic trend.

8 604 10th International Fluid Power Conference Dresden 2016 Figure 5: Influence of the oil flow rate. (a) Air release rate. (b) Bubble distribution The reason is that an additional increment of the injected air leads to the formation of bigger bubbles. These have a higher rise velocity and are released faster. This trend can be seen in Figure 6b as well, where the bubble diameter distribution is plotted for different air flow rates until a bubble diameter of. In comparison with the distributions illustrated in Figure 5b, the three histograms do not show any significant difference. Figure 6: Influence of the air flow rate. (a) Air release rate. (b) Bubble distribution 4.2. Variation of the tank design The influence of the distance between the inlet and the outlet ports has been investigated. The air release rate is plotted as a function of the oil flow rate and

9 Group M - Hydraulic Components Paper M of the air flow rate in Figure 7a and Figure 7b respectively for the first, the second and the third tank configuration. In both graphs, the second design shows a higher air release rate. This is due to the increasing distance between the inlet and the outlet pipes. The larger this distance is the more time the bubbles have to reach the free surface of the fluid. Figure 7: Influence of the distance between the inlet and the outlet ports. (a) Oil flow rate. (b) Air flow rate The usage of a separation plate has been experimentally analysed as well. The air release rate is plotted as a function of the oil flow rate and of the air flow rate in Figure 8a and Figure 8b respectively for the third and the fourth design. The fourth one shows an higher air release rate for both variations. This is due to the separation plate. Though the return and the suction pipes are placed as in the third design, the plate forces the bubbles to follow a longer path before reaching the outlet port. It increases the time that bubbles have to be released as well as encouraging coalescence phenomena due to the collection of bubbles near the plate. Finally, the results of Figure 7 and Figure 8 show that the internal design of the hydraulic reservoir strongly optimises the air release rate.

10 606 10th International Fluid Power Conference Dresden 2016 Figure 8: Influence of the separation plate. (a) Oil flow rate. (b) Air flow rate 5. Summary and outlook To improve the efficiency of fluid power systems, reservoir designs must take into account the presence of an air phase in the working fluid. The hydraulic reservoir performs the function of releasing the air to the surroundings. The challenge is to downsize the reservoir while maintaining a low air content at the suction side. For this scope, a quantitative analysis of the air release in hydraulic tanks is required. This paper presents several experiments to investigate this phenomenon. The results show that the relative air release decreases with an increase in air load and oil flow rate. Additionally, the internal tank design strongly influences the air release process. Maximising the distance between the suction and the return lines as well as the use of e separation plate optimises the air release and saves installation space. The validation of a CFD code with test results and measurements of the time-dependent air release process are currently underway. 6. Acknowledgment The authors thank the Research Association for Fluid Power of the German Engineering Federation VDMA for its financial support. Special gratitude is expressed to the participating companies and their representatives in the accompanying industrial committee for their advisory and technical support.

11 Group M - Hydraulic Components Paper M References /1/ Busch, A., Jürgen, G.: Ölbehälter Optimierung für die Zukunft. Ölhydaulik+Pneumatik Journal, Nr. 1-2, /2/ Siebert, C., Longhitano, M., Murrenhoff, H.: Einsatz von Kavitationsmodellen in ölhydraulischen Systemen. Ölhydraulik+Pneumatik Journal, Nr. 4, /3/ Haas, R., Manhartsgruber, B.: Compressibility Measurements of Hydraulic Fluids in the Low Pressure Range. FPNI-PhD Symposium, West Lafayette, USA, /4/ Kim, S.: Measurements of Effective Bulk Modulus and its Use in CFD Simulation. PhD Thesis, RWTH Aachen University, /5/ Lipphardt, P.: Untersuchung der Kompressionvorgänge bei Luft-in-Öl- Dispersionen und deren Wirkung auf das Alterungsverhalten von Druckübertragungsmedien auf Mineralölbasis. PhD Thesis, RWTH Aachen University, /6/ Schrank, K., Murrenhoff, H., Stammen, C.: Investigation of Different Methods to Measure the Entrained Air Content in Hydraulic Oils. Proceedings of the ASME/BATH Symposium on Fluid Power & Motion Control, FPMC2014, Bath, United Kingdom, /7/ Müller, L. et al.: Messverfahren und numerische Modellierung von Kavitation in einem ölhydraulischen Ventil. Ölhydraulik+Pneumatik Journal, No. 2, /8/ Longhitano, M., Murrenhoff, H.: Experimental Investigation of the Air Bubble Behaviour in stagnant mineral Oils. ASME/BATH Symposium on Fluid Power & Motion Control, FPMC2015, Chicago, Illinois, /9/ Weimann, O.: Die Abscheidung von Luftblasen aus Schmierölen durch konstruktive Maßnahmen. PhD Thesis, TU Darmstadt, /10/ Wohlers, A.: An Approach to Optimize the Design of Hydraulic Reservoirs. 14 th Scandinavian International Fluid Power Conference, Tampere, Finland, /11/ Ti, V., Lovrec, D.: Air Release and Solid Particle Sedimentation Process within a Hydraulic Reservoir. Tehnicki vjesnik 20, No. 3, pp , 2013.

12 608 10th International Fluid Power Conference Dresden 2016 /12/ Untch, J., Vollmer, T., Lang, T.: Approach for the Investigation and Evaluation of Hydraulic Tank Designs regarding Air in Oil Behaviour. 9 th International Conference of Fluid Power, Aachen, /13/ DIN : Fluid power; Hydraulic Reservoirs made of steel; dimensions, requirements, test methods; nominal capacity 63 to Norm, /14/ Franklin, R. E.: A note of the Radius Distribution Function for Microbubbles of Gas in Water, ASME Cavitation and Multiphase Flow Forum, FED-Vol. 135, pp , Nomenclature Geometrical constant - Drag coefficient - Bubble diameter mm Gravitational acceleration m/s 2 Air flow rate Nl/min Oil flow rate l/min Air release rate % Bubble Reynolds number - Air volume m 3 Bubble velocity mm/s Total volume m 3 Air fraction at the inlet port - Air fraction at the outlet port - Dynamic viscosity of the oil kg/(s m) Oil density kg/m 3 Air density kg/m 3